1. Field of the Invention
The invention pertains to a system and method for chemically coating a variety of surfaces with semiconductor materials, metals, or insulators for various applications including electronics.
2. Description of Related Art
Numerous coating processes are commonly employed in industrial applications, including electroless chemical, chemical vapor and physical vapor depositions. Physical vapor deposition is commonly used in semiconductor manufacturing applications, often employing expensive vacuum techniques in order to sustain a relatively high film growth rate. Many such processes, while performed at high temperatures (e.g., greater than 300° C.), are non-equilibrium, often resulting in non-stoichiometric proportions. Also, due to the nature of the deposition processes, the deposited films often include relatively high defect densities. In the case of semiconducting devices, such high defect levels can limit electrical performance characteristics. In semiconductor device fabrication wherein p-n junctions are formed in a partial vaccuum by depositing one film over a second film or a substrate of different conductivity type, the conventional evaporative and sputtering techniques may provide unsatisfactory film qualities. As an alternative, relatively more expensive techniques such as Chemical Vapor Deposition (CVD), Molecular Beam Epitaxy (MBE), pulsed laser deposition, and atomic layer epitaxy, are useful, especially with formation of III-V compound semiconductor materials, but satisfactory deposition processes have not been available for fabrication of thin film II-VI compound semiconductor materials.
Chemical bath deposition (CBD) is a low cost, low temperature technique which, under certain limited conditions, can provide high quality thin film growth of desired stoichiometry. However, the technique has provided low growth rates, e.g., 20 to 30 Å/minute, and the grown film thickness per bath cycle is limited to around 1000 Å rendering it non-suitable for volume manufacture. Generally, for products, such as solar diodes and light emitting diodes, film thicknesses on the order of 3,000 to 20,000 Å are needed. With traditional chemical bath deposition techniques, efforts to increase film growth rates typically result in degradation of film quality. Thicker films of higher quality, however, can be attained by cyclic deposition of stacked relatively thin layers. This, of course, is time consuming and generally can be an expensive endeavor, wherein relatively expensive chemicals are wasted due to deposition of film materials on unwanted surfaces such as the equipment.
Conventional CBD is carried out with the source solution held at slightly elevated temperature, typically ˜85° C. [see, for example, Oladeji and Chow, “Optimization of Chemical Bath Deposited Cadmium Sulfide Thin Films,” J. Electrochem. Soc. 144 (7): 2342-46 (1997)]. In this situation, heterogeneous nucleation of the CdS film on the substrate must compete with homogeneous nucleation of colloidal CdS particles within the stirred reactant solution. Particulates represent not only a waste of reagents but also a source of defects in the deposited film.
A process for depositing ultra-thin semiconductors is taught by McCandless et al. in U.S. Pat. No. 6,537,845. The process uses a premixed liquid containing Group IIB and VIA ionic species and a complexing agent. The solution is applied to a substrate heated to a temperature from 55 to 95° C., forming an ultra-thin (100-500 Å) coating. For thicker coatings, the process can be repeated. The process taught in '845 suffers from several noteworthy shortcomings. First, using a single complexing agent (generally taught to be NH4OH) prevents adequate process control: at a low concentration the solution is so unstable that unwanted homogeneous nucleation can occur, whereas at high concentration the activation energy required to form the film becomes so high that the claimed substrate temperature may not be able to overcome it to cause a film growth. Second, the thickness of film that can be grown in a single step is very small, so to grow a film of 0.1 μm or greater, multiple cycles are needed and this will tend to introduce greater concentrations of defects. This will also make the process cumbersome and less manufacturing friendly.
Films have also been prepared using the flowing liquid film process as taught by Ito et al. [Preparation of ZnO thin films using the flowing liquid film method, Thin Solid Films 286: 35-6 (1996)]. The deposition process involved the reaction of zinc chloride and urea at 70° C. according to the reaction:ZnCl2+NH2CONH2+2H2O→ZnO+2NH4Cl+CO2.The use of NH2CONH2 as a single ligand will result in excessive homogenous reaction, and ZnO film growth takes place mostly by particulate adsorption. ZnO so formed is less transparent and of little or no practical use. Furthermore, the side of the substrate next to the incoming solution gets the full dose of the solution, hence high growth rate, whereas the side of the substrate at solution exit receives the least dose, hence low growth rate. As a result, the film non-uniformity will increase with increasing substrate size. Since all the flowing solution sees the heat provided by the substrate, film growth will take place on the substrate plus any part of the system on contact with the hot growth solution, leading to material waste. If the particulate generation happens in the enclosed chamber, these could be trapped and get adsorbed unto the substrate, leading to poor quality film.
In U.S. Pat. App. Pub. 2003/0181040 as taught by Ivanov et al., films have also been prepared on semiconductor substrates using a sealed chamber filled with growth solution maintained at high pressure about 2 atmosphere and temperature at 0 to 25% below the boiling point of the solution. The use of high pressure is needed to keep the rigid substrate in place, and the cost associated with this is not trivial. The relatively large volume of solution in the chamber is heated completely by the heater located outside the chamber or in the substrate holder or both. This arrangement forces the hot growth solution to be in contact with the substrate and substantial part of the chamber system, and results in wasteful film growth on unwanted areas. The latter will also increase the cost of keeping the chamber clean to prevent particle build up. Subjecting a relatively large volume of growth solution in chamber at high pressure and temperature will not only encourage the desired heterogeneous reaction responsible for film growth but also substantially increase the unwanted homogenous reaction; the latter will lead to fast depletion and waste of material, and this will in turn increase the production cost.
These facts were recognized by Ivanov, who through U.S. Pat. No. 7,235,483, attempted to minimize the material waste, by having the substrate heated and cooled instantaneously, and using the high temperature only during the growth regime when it is needed, especially during the bulk film growth step. For Cu interconnect, the focus of that work, where 500 Å or less Cu metal seed and CoWP Cu capping layer are needed, the bulk deposition step may be short enough to prevent quick depletion of chemicals. However, for semiconductor film deposition needed for solar cells and other optoelectronics applications where the required film thickness is more than 1000 Å, the high temperature bulk deposition step time will be longer and the instantaneous heating and cooling will not help. Generally, in electroless semiconductor film deposition catalyzed substrate surface is not required. One would therefore expect the growth rate on the substrate as well as other parts of the chamber system in contact with the hot solution to be about same; the consequence of this is substantial material waste. This will make the chamber cleanliness even a much bigger challenge. The operating cost and waste management in using this system for semiconductor film manufacturing will be prohibitive. It should also be noted that this approach is only applicable to rigid substrates; flexible substrates cannot be used.